Abstract
Background: Breast cancer is the most common cancer in women, and it is important to identify models that can accurately predict mortality in patient with this cancer. The aim of present study was to use the elastic net regression and artificial neural network models in diagnosing and predicting factors affecting breast cancer mortality.
Study design: A cross-sectional study.
Methods: Data of 2836 people with breast cancer during the years 2014-2018 were analyzed. Information was registered in the cancer registration system of Kerman University of Medical Sciences. Death status was considered as the dependent variable, while age, morphology, tumor differentiation, residence status, and residence place were considered as independent variables. Sensitivity, specificity, accuracy, area under the ROC curve (AUC), precision and F1-score were used to compare the models.
Results: Based on the test set: the elastic net regression determined factors affecting the breast cancer mortality, (with sensitivity (0.631), specificity (0.814), AUC (0.629), accuracy (0.792) precision (0.318) and F1-score (0.42)), and artificial neural network (with sensitivity (0.66), specificity (0.748), AUC (0.704), accuracy (0.738) precision (0.265) and F1-score (0.37)) did so.
Conclusion: The sensitivity and AUC of the artificial neural network model were higher than those of the elastic net regression, but the specificity, accuracy, precision and F1-score of the elastic net were higher than those of the artificial neural network. According to the purpose of the study, two models can be used simultaneously. Based on the results of models, morphology, tumor differentiation and age have a greater effect on death.